Capitalisn’t: The Case That Could End Super PACs

Episode Transcript

Lawrence Lessig: I feel like Citizens United is the most misunderstood decision because Citizens United was not saying you couldn't police coordination, but the failure to police coordination is a failure of the FEC. It's a failure of the prosecutors. It's a failure of the system to actually live up to the policing that the Supreme Court has allowed.

Bethany McLean: We all know that the 2010 Citizens United Supreme Court decision is the root of all evil, don't we?

Luigi Zingales: It equated corporation to people, it equated money to speech, and it created the so-called super PACs, political groups that can raise and spend unlimited amount of money to help or hurt political candidates.

Bethany McLean: Basically, Citizens United is responsible for the state of US politics.

Luigi Zingales: Except that's completely wrong.

Lawrence Lessig: When you say, "Why do we have super PACs?" You're going to say, "It's because of a Supreme Court decision called Citizens United." That's just false.

Bethany McLean: That's Harvard Law professor Larry Lessig at this year's Stigler Center annual conference.

Lawrence Lessig: We have super PACs because of a lower federal court decision interpreting Citizens United three months after it was decided, a case called SpeechNow. The upshot of this is SpeechNow is a mistake. Under Citizens United, SpeechNow is a mistake, and therefore fundamentally vulnerable.

Luigi Zingales: You often hear a lot of interesting things at an academic conference, but rarely something you think will change anything. This time we did.

Lawrence Lessig: People like my friend Bernie Sanders make a million speeches about how if we believe in democracy, we must overturn Citizens United. Overturning Citizens United is technically impossible. It just cannot happen. There is zero chance the United States Supreme Court is going to overturn Citizens United. The idea that you would rally people to an impossible cause is absurd, but the idea that the Court would recognize that SpeechNow doesn't actually follow from Citizens United is not impossible. I think it's quite possible.

Bethany McLean: After that talk, we knew we had to get Larry on the podcast to tell us more. With election season in full force, that's exactly what we're doing on this episode.

Luigi Zingales: Larry's argument is subtle, as many legal arguments are. Citizens United simply stated that independent expenditure don't give the appearance of corruption, and as such, they cannot be limited. As it is my constitutional right to express support for Larry Lessig for president, actually, Larry did run for president-

Bethany McLean: Did he?

Luigi Zingales: Yes, he did.

[laughter]

Bethany McLean: I didn't know that. Wow.

Luigi Zingales: -it is my constitutional right to spend $100 million in supporting his candidacy as long as I do so independently. I do not coordinate with him and his campaign.

Bethany McLean: So far, there is nothing about contributions, only expenditures. This is where the SpeechNow decision comes in. It ruled that limits on individual contributions to independent expenditure only political groups violate the First Amendment. This is what created super PACs.

Luigi Zingales: Larry claims that individual contribution can be and are coordinated with political candidates. Candidates even fundraise for super PACs. There can be the possibility of a quid pro quo, and as such, this contribution can be regulated.

Bethany McLean: In November 2025, 75% of Maine voters approved a law limiting contributions to super PACs to $5,000. A local judge blocked the enforcement of the law, claiming that the Supreme Court's 2010 decision in Citizens United clearly prohibits any limits on contributions.

Luigi Zingales: This is where Larry comes in. Citizens United is about expenditures, not contribution. He appealed the judicial decision. The appeal was argued this July before the First Circuit, and in September, the First Circuit asked for supplemental arguments. The decision is due any day now. If Larry wins, the Supreme Court must intervene because there would be an inconsistency in how the law is interpreted across the various districts.

Bethany McLean: This decision can transform the 2028 electoral campaign. Therefore, it's worth hearing Larry's arguments just for that, but it's more. This suit is part of a lifelong battle Lessig is waging against institutional corruption. Larry, before we talk to super PACs, I wanted to ask you to define the word that underlies a lot of what you've worked on for the last 20 years. What is corruption?

Lawrence Lessig: The most obvious sense of corruption is bribery, when an official takes money in exchange for doing some favor, and that's regulated, and it's prohibited, but I don't think that's actually the most important kind of corruption. The most important kind of corruption we could call dependence corruption, corruption that develops when you have a dependence on a source of income that creates a conflict with your underlying purpose in doing what you're trying to do.

The framers of our Constitution, Madison, told us we would have a Congress that would be "dependent on the people alone," not the rich more than the poor. We've developed a system for funding campaigns that makes candidates dependent on the tiniest fraction of the 1%. They know if they don't make that tiny fraction of the 1% happy, they're not going to have the money they need to be able to run their campaigns. That's a dependence that's inconsistent with the dependence on the people alone, and therefore that, in my view, is also a kind of corruption.

Luigi Zingales: Now, among the many problems of American democracy, you think that there is a Tweedism problem. Can you explain to our listener what is this Tweedism problem?

Lawrence Lessig: Yes. Boss Tweed, who was the head of Tammany Hall in the 19th century, used to say, "I don't care who does the electing as long as I do the nominating." We've created a similar kind of Tweedism with money in politics because what we know is that if you want to be a candidate running for Congress, what the Democratic Party will tell you, what the Republican Party will tell you, is the single qualification to being a candidate to run for Congress is your ability to raise literally millions of dollars. If it's a swing district, it can be tens of millions of dollars.

That means that the funders, in a sense, become Boss Tweed. They're the ones who are essentially controlling the nomination, the nomination of who's even allowed to run. Once you have candidates, all of whom are responsive to the money, then they don't need to worry about who wins, the Democrat or Republican. Doesn't matter who wins, as long as the dependence is there, and that dependence is what creates the corruption inside the system.

Luigi Zingales: There is another point that I heard you make: when basically the main business of a politician today is not to kiss babies and figure out what the problems in the world are, it's to raise money. They raise money all the time, and this selects a kind of people that like to kiss up, to use a technical term. As a result, we have a bunch of congressmen and senators who cannot stand up to anybody as a result of the selection process. Money in politics is not a problem of Republican or Democrat, so much so that the base of both parties hates this. It is a problem of representation.

Lawrence Lessig: Yes, it's a great point. If you think about the dynamic of being a member of Congress right now, the one thing you know they do is spend two to four to six hours every day sucking up to people with power. What does that do to somebody psychologically? Who's going to do it? Only the sort of people who are willing to suck up like that. Even if you're a person willing to suck up like that, on day 1, you're going to be different from how you are on day 150, because you're practiced in sucking up like that.

Is it any surprise we have a Congress filled with sycophants, that the Congress is filled with people who don't have the courage to do what they know they believe in, what they know is right? That is, I think, a deep, deep problem caused by the corruption of money.

Bethany McLean: Is that the answer to the argument in the political science literature? A large body of research shows that money in politics actually has little effect on policy. Is what that gets wrong is that maybe that's measuring a micro effect on policy, where the macro effect is that the very people who are there are there because of the money?

Lawrence Lessig: Yes, absolutely. I feel like that's work is like somebody who takes a drone and flies it over the Catholic Church and sees no pedophiles anywhere, and then says, "There's no problem with pedophilia in the Catholic Church." Okay. Using the tools you're using, you're not going to see the problem. In this context, the way money has its effect is very subtle in Congress. For example, it determines whether a bill gets voted on or not, whether it comes out of committee or not. Now, how are you going to see that in the tracking which this money is doing?

Bethany McLean: I think there's a tendency to use the word "billionaires" as shorthand for agreement. I don't know a lot of billionaires, but the ones I do know disagree with each other radically. I know billionaires who support Trump. I know billionaires who support Kamala Harris. I know billionaires who support tax increases on the wealthy. I know those who don't. Why could you not make an argument that the increase in billionaire spending simply cancels each other out given they don't move as a class? There's no uniformity there. There's actually a fair amount of ideological difference.

Lawrence Lessig: Okay, but those are two different claims. Yes, they cancel each other out ideologically, but that doesn't show there's no uniformity on class. Billionaires don't worry about dental health insurance, but there are a lot of ordinary people who worry about the fact that they don't have dental insurance. They can't get it because they can't afford it, or can't get health insurance. There is a class difference. There just isn't necessarily an ideological difference.

Bethany McLean: Just to push on this for a second, your argument is that it's the issues that aren't getting considered and that aren't raised to the level of importance rather than the issues that are and are perhaps canceled out by the competing ideologies within the billionaire class?

Lawrence Lessig: Exactly. It's what issues can't get considered. What issues can you not raise or Congress can't begin to consider because they know that some billionaire's going to come in and dump a bunch of money against them if they raise that particular issue. Money given can be just as effective as the threat of money that you will give. If I'm a PAC and I'm credibly able to say, "I could give to the Republican or I could give to the Democrat, and if you don't follow me, then I'm going to give to the Democrat," then that's just like giving to the Democrat from the standpoint of the effect it has on the person you're threatening.

Look at the crypto super PACs that were so successful in 2024. They just took out a couple prominent members of Congress, and they promised support for a couple prominent members. Sherrod Brown's taken out, and a couple others are taken out. What that does is it creates an environment of fear in Congress, and congressmen are pretty cowardly, basically. They're not willing to risk much. The very idea that there's a super PAC out there that's going to take them on is enough to force them to basically flip crypto policy in one election cycle. The same thing's going to happen, I fear, with the AI super PACs that are standing themselves up to be involved in this election.

Bethany McLean: Another foundation of your thinking that I wanted to push on a bit before we come back to Citizens United is this idea of transparency. I think I had always bought into the conventional notion, the Brandeisian one, that sunlight is the best disinfectant. In reading your work, when you make the point at times bluntly that transparency solves nothing, I think that is an entirely different way of thinking. Most of us believe, pull it out into the light of day and it will all be fine.

Lawrence Lessig: Yes, that's right. It's striking to me how transparency thought just bends how people think. There's this great incident where Mike Johnson is asked about his obsessive desire to impeach Joe Biden for Joe Biden's corruption, but his complete unwillingness to question the president about the president's corruption. He says, "The difference is Joe Biden tried to hide his corruption, but the president is completely transparent about everything he does," as if that makes it better, when obviously it makes it worse.

If you're transparent about your kleptocracy, then it just creates a kleptocratic regime. Everybody knows exactly what they need to do. If you at least have the shame to hide it, then there's some ambiguity about whether it's a kleptocratic regime. This is, I think, a product of our belief that transparency, über alles, that is a perfect solve for any problem. The reason it isn't is that it might convey information to the public that convinces the public that they're irrelevant. They don't matter.

If we only knew who the money came from, then we wouldn't have a corrupt system. No, we would have a more corrupt system because more people would look at this and say, "Oh, I see why they're not willing to do X, Y, and Z, because now I know it's because this super PAC was funded by that super billionaire," whatever. If all you've got is transparency, then you've got a transparently corrupt system. What you need is a non-corrupt system that you're transparent about, and then people have a reason to trust it again.

Luigi Zingales: One thing that most people forget is how much worse things have become in the last few years. Reagan, when he ran for the reelection campaign in 1984, did zero fundraising events. Obama, in 2012, did 220 fundraising events. Now, if you are the sitting president and you meet with people, the opportunity for potential corruption are very, very large, and 220 means almost 1 a day during the campaign. Now, everybody blames the infamous Citizens United decision, but you, Larry, say that's wrong. Can you explain us why?

Lawrence Lessig: Yes, but I want to put an underline under some of what you just said. The first point is just to recognize every president between Richard Nixon and Barack Obama was elected with public funding. Ronald Reagan would never have been president. There would never have been any money to support him challenging a sitting president. It was only because there was presidential public funding that he could rely on to be able to raise the money necessary to challenge Ford. It's astonishing to recognize that we had this period of presidencies where the president didn't need to worry about money, didn't need to worry about how he was raising money, and didn't try to leverage his power to raise more money. That's point number one.

Point number two is you're exactly right. This has been an explosion in the size of the money, but I want to emphasize, even more troubling is the concentration of funders. I don't know how much money it should cost to run presidential or congressional campaigns. We spend multiples of the amount we'd spend electing a president advertising soap every year. Maybe there should be more money in politics. I don't know. What I do know is when you concentrate power in just a very few funders, that increases the opportunity for them to leverage that power into the distortion of policy. In 2010, 0.3% of outside spending was from billionaires, and in 2024, that number was 19%, so it's a 63-fold increase.

Now, in 2010, the reason this is the baseline we all measure from, the Supreme Court decided Citizens United. Citizens United said, following a decision in 1976, Buckley v. Valeo, that if a corporation spends its money independently of a campaign, the government has no legitimate interest in limiting the size of that spending. The reason the government would have no legitimate interest is, according to Buckley, the only justification the government could have for limiting political speech is the justification that it creates a risk of something called quid pro quo corruption, that kind of bribery corruption we were talking about before. The reasoning was, if it's independent, then there's no such risk.

Now, the part of that story which people forget, because who cares to know? Only 10 people in the world want to be campaign finance experts. The part they forget is that the law actually has a mechanism that says if it turns out that your independent expenditure, in fact, was coordinated, like if you talked about where it should be spent or how much needed to be spent or what's the right targeting audience with the campaign, then the law considered that expenditure a contribution to the campaign.

If Elon Musk comes and spends $10 million to elect people to the Supreme Court of Wisconsin, and it turns out that spending was coordinated, then that $10 million expenditure is converted into a $10 million contribution, and that's about $9,999,000 more than he's allowed to contribute. It's a violation of the law. That's important because the step that led to super PACs was a step made by a lower federal court two months after Citizens United.

The DC Circuit, in a case called SpeechNow v. FEC, decides that, if the spending is independent and therefore creates no risk of quid pro quo corruption, it follows as a matter of logic that a contribution to a committee that's going to spend that money also cannot create a risk of quid pro quo corruption. It went from the committee is independent to the donor is necessarily independent if the committee is independent. That's just ridiculously, obviously, just not true.

The committee could be independent in the sense that it doesn't talk to the candidate, but a donor could go talk to the candidate and say, "What do you need?" The candidate could say, "I need some more money in my super PAC." The donor says, "Fine. I'll go give money to the super PAC. Tell me where they should spend it or how they should spend it." The candidate and the donor make out this deal, and then the donor goes and says to the super PAC, "Here's $10 million, and I want you to spend it in the following way." This was decided in 2010.

In 2015, it was a perfect example of exactly this kind of corruption. Robert Menendez, Democrat from New Jersey, was indicted in 2015. The allegation was that in exchange for Menendez promising favors to some Florida businessman, the Florida businessman would give money to Menendez's super PAC. There it is, a quid pro quo involving a contribution to a super PAC, the very thing the DC Circuit said could not happen, as a matter of logic, could not happen.

Indeed, his lawyers had the chutzpah to file a motion to dismiss. The motion said, "The thing you've indicted us for, the DC Circuit says cannot happen. Therefore, we must be innocent. Therefore, dismiss the indictment against us," which, of course, the court just chuckled. The point is, it showed the logical mistake at the core of SpeechNow, and it's SpeechNow which created super PACs, not Citizens United.

Luigi Zingales: I love your argument and the difference between contribution and expenditures. If we look from a practical point of view, and I know that the law is not always practical, but from a practical point of view, the independence of expenditure is a fig leaf, it's fake, in a sense. The PAC at the headquarters in the same building as the campaign.

They could coordinate every time they go pee.

Not only that, there is a guy at the University of Chicago that shows that the candidates in their websites, they have a red box to say, "This is a message that you, please, Super PAC, should advertise." It says, "If this is not coordination, what is coordination?" Why nobody has raised the issue? I cannot believe that in 15 years nobody coordinated, but I know that in 15 years nobody got caught, so it must be that nobody's looking, right?

Lawrence Lessig: Yes, that's exactly the point, Luigi. Sometimes I feel sorry for the Supreme Court, which is weird because they've done so many things not to be sorry about. I feel like Citizens United is the most misunderstood decision because Citizens United was not saying you couldn't police coordination. It was not saying there wasn't a role for the FEC to make sure that there wasn't actually coordination between the expenditure and the candidate.

What it was saying was, so long as there wasn't coordination, there was no justification for limiting the expenditure. The failure to police coordination is a failure of the FEC, it's a failure of the prosecutors, it's a failure of the system to actually live up to the policing that the Supreme Court has allowed. The Supreme Court's perfectly fine saying that if, in fact, you've coordinated, if you've done no more than simply talk to the campaign and discuss with the campaign, where would this expenditure make most sense? That's all coordination requires.

If you've done even just that, the Supreme Court's perfectly happy to say, "Okay, now your expenditure is going to be treated as a contribution, and therefore as a criminal contribution," because a $10 million contribution is way above the limit. They're willing to allow the regulation, but they require the enforcement, and there is no enforcement because basically everybody in the system, Democrats and Republicans, likes the system as it is, or they don't like it, but they're not willing to change it. Therefore, they go along with a system where there's effectively no enforcement of the existing regime, and people blame the Supreme Court. Again, the Court didn't create that problem.

The Court created an infrastructure where that problem could be addressed, and I think it should be addressed. Indeed, I think in the next election, we should go back and look at the so-called independent expenditures that were made in this last election, and we should determine were they actually independent? Was there any coordination? Because if there was coordination, then there needs to be some kind of enforcement action. The best example of this is the President himself.

The President was asked in an interview about the Trump super PACs. The interviewer was very careful to say, "The super PACs that are supporting you." She originally said, "That you control," but then she backed away. She said, "No, no, the super PACs that are supporting you." Trump said, "No, no, no, I control them. They are my super PACs. They are mine." It's like, you just blew up your super PAC because you've just acknowledged that it's not independent.

Therefore, anybody it spends money against should have the opportunity to bring a lawsuit saying that you can't spend more than $5,000 against me because it's a coordinated expenditure. It's not an independent expenditure. The world is filled with people who think there's no rules here. Certainly, the President does, and so therefore he feels no hesitation at admitting the fig leaf is not even a fig leaf.

The challenge is the FEC doesn't function anymore. It doesn't have a quorum. It doesn't have sufficient ability to even make a decision. That's why this whole enforcement mechanism is broken right now, and it's intentionally broken. They don't want to create the opportunity to enforce it, so basically they've shut the agency down.

Bethany McLean: I want to go back to what happened in Maine because it's a little bit confusing, and it seems to me that it is both good news and bad news. Bad news in the sense that it wasn't a victory, but good news in the sense that what the judge said actually opened the door quite substantially. Can you explain that?

Lawrence Lessig: Yes. The initiative was on the 2024 ballot. It won with 74.9% of the vote, 600,000 votes. More votes were for this initiative than ever was cast for any candidate or initiative in the history of Maine. Democrats, MAGA, everybody hates money in politics. Indeed, when I gathered signatures for the initiative in 2023 on Election Day, first guy to come up to me said, "I'm here gathering signatures to get Donald Trump on the ballot. What are you here for?" I showed him my initiative. He read it. He said, "Hell yes," and he signed it. He said, "I'm going to get everybody who signs my initiative to come over and sign yours too," which he did for the hours that he was there. First list of my signatures are a whole bunch of MAGA Trumpers.

The point is, it passes overwhelmingly. Two super PACs bring a lawsuit, one month after it's passed, challenging the initiative. Then a district court judge issues an opinion striking the initiative down. What distinguished her opinion from the opinion of every other court before is that while every other court before repeated the mantra, a contribution to an independent political action committee creates no risk of quid pro quo corruption, she acknowledged that a contribution to an independent political action committee can create the risk of quid pro quo corruption.

Now, at that point, she should have said, therefore, of course, the people of Maine have the right to regulate it because the Supreme Court has again and again and again and again and again said, "If you're addressing corruption, quid pro quo corruption, the state has a legitimate interest in regulating." Instead, she crafted a whole new First Amendment theory that shows that even though there was a risk of quid pro quo corruption, there was nothing the people of Maine could do about it.

That's what tees up the opportunity for the Court of Appeals because that's the most extreme opinion we've ever seen in a federal court. There's never been a federal court that said, even to address what's acknowledged to be corruption, there's nothing you can do.

We think that the argument which our lawyer, Neal Katyal, made in alliance with the Attorney General of Maine was to say that under the reasoning of Citizens United, under the logic of Citizens United, embrace Citizens United. Under that reasoning, you can regulate contributions because, as the district court said, they create a risk of quid pro quo corruption, even though you can't regulate the expenditures, because so long as the law is enforced, the expenditures can't create a risk of quid pro quo corruption.

Luigi Zingales: Now, I heard you saying that besides this line of argument, in case the Supreme Court goes originalist on you, you have an originalist argument. Can you make that argument too?

Lawrence Lessig: Yes. This is a kind of embarrassment of originalism because originalism has made its mark in a whole bunch of areas, but it's never yet been applied in the context of the First Amendment, free speech clause. Justice Thomas has been trying to get it applied in the context of the First Amendment, free speech clause, by getting New York Times v. Sullivan overturned. That was the case that created an immunity for reporters and newspapers if their errors are just negligence, not intentional.

What Thomas says is we ought to decide those cases by asking, what would the framers have said? What would the decision of a court have been in 1791? He wants to apply the originalist methodology to the First Amendment because, as he said, talking about New York Times v. Sullivan, that decision is policymaking dressed up as constitutional law.

When I read that, I thought, "Have you ever read Buckley v. Valeo, Justice?" There could be no opinion that's more clearly policymaking dressed up as constitutional law, because the court, completely out of whole cloth, kind of Zeus-like standing on Mount Olympus, just declares a whole principle about how you're going to police the ability of a legislature to police the integrity of their own relation. It was never before was that crafted, never before was this articulated, they just made it up.

The point is, if you were consistent in your application of originalism, originalism essentially means Roe v. Wade is wrong, originalism means everybody gets a right to bear arms. If those are your principles under originalism, then you ought to apply originalism in the context of campaign finance as well. What we demonstrate through expert briefs filed by some of the very best historians, including Jonathan Gienapp and Jack Rakove, is that the original meaning of the First Amendment plainly would have allowed the regulation of the size of contributions to independent political action committees.

Indeed, our champion on this is Josh Hawley, the senator from Missouri, because Josh Hawley, in 2023, introduced a bill in Congress to overturn Citizens United. When he introduced that bill, what he said was, "Everybody knows that the original meaning of the First Amendment would never have restricted the ability of Congress to control how corporations spend their money," period. That's absolutely true. The point is, if we can get the originalists to recognize what their principles require, I don't think we're going to get Justice Thomas. I don't think we would get Justice Alito.

I think justices like Amy Coney Barrett or Justice Gorsuch are trying to be Scalia wannabes. I say that with affection. I worked for Scalia. What I mean by that is somebody who's trying to apply their principles in a principled way without it seeming that what they're doing is just politics. I think if we can get them to think about this, they will recognize why their principles should steer them in the direction of upholding this initiative. Even if it doesn't lead them to overturn Buckley, at least it might lead them to be more receptive to the quid pro quo argument that's at the core of why SpeechNow was wrong.

Bethany McLean: I wanted to, in the last minutes of our conversation, turn back to where a good chunk of your early intellectual life and how that comes full circle with what you're talking about now, which is the dangers posed by AI. As we head into this world where AI companies and people behind it can exert unlimited spending, what that might mean, and if that increases the onus to get this fixed sooner rather than later. I'm also curious what you're hearing from your Harvard students as you teach them.

Lawrence Lessig: The Harvard students are terrified because they realize, law students realize, the law's about to change dramatically. I think for good, in a good way. I think 70% of lawyers are going to disappear, and that's going to be great because that's going to radically lower the cost of law, and I think radically improve the quality of law. People are so anxious. They want to police the use of AI to write briefs. I say to them, "Have you ever read briefs filed by lawyers even in the [chuckles] Supreme Court? They're pretty terrible.

They're really bad." If you compare them to a fable-written brief, the machines are pretty good right now. I think it's inevitable.

I also think, even though I think there are great things AI is doing, I've spent a long time, the last three years now, focused intensely on this AI risk question, a regular call with some of the leading policymakers, but also AI scientists about this. I'm convinced that we've got to be careful, and I'm also convinced we don't have a government now that has anything close to the understanding of the risk and willingness to address it.

The question is whether we could have a government that had the ability to do something about it. I fear that these AI companies are following the playbook of the crypto super PACs, which follow the playbook of AIPAC, which is basically to create an aggressive super PAC that signals exactly what lines you can't cross and creates a Congress that's terrorized and unwilling to do anything to set up an infrastructure to regulate even risks, which insiders say creates a substantial chance of existential events.

Now, you can be skeptical about what they're saying. I'm not, because I think these are, especially the people who are not inside the companies, but who are the most informed, scientists like Yoshua Bengio and Geoffrey Hinton. These are people who understand this technology better than any of us, and they're the ones that are most anxious about the risk, but I understand people who worry that maybe the companies are doing this for strategic reasons. Maybe OpenAI wants to be acquired by the government because they realize they don't have a business model and they have too much debt, so this is a great way to escape the problem they've created. That's all fair skepticism.

We need a government that's free to lead. We need a government that's capable of making the right decision based on the evidence. If we have a government where Congress is dependent on not being attacked by these mega super PACs, then we don't have that government. That's why I think at this moment especially, we could end super PACs by 2028, which is possible, depending on how quickly the First Circuit decides. That would be an enormous freedom for that election to be free of the power of the super PAC money.

Luigi Zingales: I don't know to what extent this catastrophe is correct or not. Even if it is correct, why isn't the solution to improve the liabilities of this company? After all, if I am designing a self-driving car, if I put enough liability when I kill people, the company will be very careful in deploying it. I fear that there is not enough liability in this particular case. My fear is that the first thing these companies will obtain from the government is a reductionist liability that will go exactly in the opposite direction. I would be even in favor of a radical thing to say, "Look, tort claimants should become senior in bankruptcy." A lot of the financial data center will disappear overnight, if we were to put that.

Lawrence Lessig: I completely agree, but I would go one step further. I would say we got to clarify the liability, and we have to impose mandatory insurance because the liability is ex post. Most of these people are not thinking seven years down the line, which is exactly how long it would take, even after a major event, to get any liability that actually would matter, and the companies would be bankrupt long before that.

The point is, if you enforce ex ante insurance mandates and companies were making a decision about which technology to deploy based on the insurance rates, they're going to have to pay for that. An insurance company auditing saying, "Do you really need to enable agentic AI in this context? Why isn't non-agentic AI perfectly just as capable for what you're trying to do?" It would internalize the risk judgments, and those risk judgments would be priced by the market, so long as you got the government to stand up an insurance market here.

Now, it's not a perfect insurance market because we don't have history, and these are not uncorrelated risks, and there's all sorts of weakness to how it would do. The point is, it's better than nothing. If it were structured in a way to get information into the business decisions about how you deploy these technologies, we could begin to reduce the risks systematically. Not the existential risk, but the systematic risk, which I think we're seeing manifest right now with these agents. We got to internalize that risk, and I think mandatory insurance is a cheap way to do that.

Bethany McLean: To keep going in this direction, that is obviously the worry that companies are using this as a way to get regulation in place that will somehow benefit them. That even without super PACs, that the regulatory apparatus, because of the voice of the powerful in the setting of regulations, that the regulations will be set in a way, and it will be very surreptitious how it is, but it will actually serve to preserve the incumbents and protect them. What would be the top three things you would do to make sure that regulation that is put in place to protect us all from the catastrophic effects of AI actually does that, rather than just cementing the incumbent hold?

Lawrence Lessig: The first thing I would do is I would say, let's make explicit that companies at the very forefront are allowed to agree to slow down for safety reasons. They say that they're worried about antitrust liability. Indeed, there's already a lawsuit that's been filed claiming that their coordination violates antitrust liability. I don't think there should be antitrust liability if it's an explicitly crafted agreement around safety for the very frontier labs. I think I would do that, number one.

Number two, the Rubicon that the safety experts say we cannot cross is recursive self-improvement. Now, that recursive self-improvement technology is extremely hard. There's only probably two, maybe three models that are close to it right now. If you have a regulation that says you can't deploy RSI until you have an FDA-style safety clearance, again, that's not protecting these companies against competitors. That's burdening these companies, and it wouldn't burden their competitors because the companies far below them are not doing anything close to anything like RSI.

Number three, I would do this mandatory insurance, and mandatory insurance is across the board. It's obviously costly, and so companies on the margin would be burdened by it, but it's not necessarily something that's going to be protecting the incumbents. Number four, I would create explicit immunity for AI safety whistleblowers. The problem with whistleblowing in this context is that typically the whistleblower statutes protect your ability to report crimes, but reporting AI safety concerns is not reporting a crime. You have no automatic immunity, no immunity from your obligation not to reveal confidential information.

That immunity ought to be secured so that we have a way in to know exactly what's happening. If you report to a safety institute or you report to the government that they're not being honest about what's going on, that would give us a way to make sure that we know that what's going on is not unsafe. I think those four changes are not designed to protect those companies. They wouldn't, in fact, protect the most powerful of those companies, but they could be an important step in the direction of getting some safety deployed.

Luigi Zingales: By and large, many of the representatives today respond to money, and money is all in one direction. The moment they will respond to people, we see that on issues that people are very sensitive about, like data center, they are willing to deviate from money because there is a strong popular demand. I think democracy works because the will of the people is transmitted to the representative, and the super PACs are muting that channel, and by muting the channel, makes the democracy not work.

Lawrence Lessig: That's a great way to put it. It's exactly right.

Luigi Zingales: In a sense, it's inevitable that powerful companies like AI companies will have some say. I think what they are trying to do is try to rebalance a bit the imbalance. I don't think that, even Larry, please feel free to speak up, but I don't think you think that we'll be able to put a perfect level playing field.

Lawrence Lessig: No, no.

Luigi Zingales: There's a big difference.

Lawrence Lessig: Nobody should want a perfectly level playing field. That's a disaster. That kind of equality is not the objective. The objective is to remove a corrupting dependence inside of the system. If you remove that corrupting dependence, there'll be lots of inequality. There'll be lots of people who get more money because they're more popular or they have a way to persuade people to donate more money. That's fine. It's the dependence that's the problem.

Bethany McLean: I wasn't clear what you meant, and maybe it'd be helpful if you define that. What do you mean by not wanting equality? What do you mean by a level playing field then?

Lawrence Lessig: Some people say that the objective of campaign finance regulation should be to equalize speech in the marketplace, to make it so that all voices have the same power. I don't think that's the objective of campaign finance regulation. Even if you could achieve it, I don't think that's the objective. I don't think you can achieve it because I don't even know what it means, really.

What I do think we should do is to make sure that representatives are not focused exclusively in one direction because they know that one direction is the thing that's going to make it so that they can't get reelected. It's about removing this corrupting dependence, recognizing that even without the corrupting dependence, you'll have lots of super popular people who will have lots more money than other people, and that's totally fine.

Luigi Zingales: Very last thing, what are the steps if you win at the appeal court? What is the next step?

Lawrence Lessig: The ideal way that we win is that we lose and win. There's two issues in our case. One's about the disclosure requirement. One's about whether you can limit the size of contributions. If we lost on the disclosure requirement and won on the limiting the size of contributions, we would have the right to file cert to the Supreme Court. If that's what happened, we would immediately file cert to the Supreme Court complaining about the loss on the disclosure requirement, but that would then bring it up to consider both of those things at the same time. We would control the timing of that. That's the important point.

If we win on both issues, the other side may well think strategically about slowing the process down so that there's no chance that the court would have an opportunity to really resolve this before the 2028 election, which I know a lot of big money will be keen to make sure happens. That's why it's hard to predict exactly what the next steps are until we see the character of the decision the court makes.

If we lose, fully lose, then we'll file for cert, no doubt. There might be a low chance the court will grant it because there will be no split for it to review, but I think in general, we've done a good job in making it clear to people why there's a mistake here, so maybe the court would take it up. That would again give us the chance to control the timing. If we can control the timing, we're going to file as quickly as we possibly can. If we get a chance to file before January, maybe the first couple weeks of January, then the court could hear the case this term, which means they could decide the case by July of 2027.

Luigi Zingales: Bethany, do you appreciate the argument better after talking to him?

Bethany McLean: [chuckles] I do. I appreciate the argument, and I appreciate the importance of it. I guess I'm troubled both that this happened narrowly speaking and that it happened broadly speaking, because it shows how really stupid things can get enshrined in our law because people make mistakes. It's not a terrible thing that people make mistakes. I make plenty of mistakes. We could all wish we don't make mistakes, but we do. It's a really terrible thing that, in the law, they get set in stone, and they become almost impossible to deviate from, even when the underlying logic is clearly flawed, right?

Luigi Zingales: Yes, except that my Italian cynicism is coming up. Remember Winkler's episode? What was a Supreme Court reporter crafted the summary of judicial decision has so much impact in subsequent decision, but only in one direction. Mistakes happen, but they happen only in one direction, so you wonder if they are really mistakes.

Bethany McLean: That's what I was trying to get at with my question for him. Also, in Winkler, that this really fundamental issue, are corporations people-- what are they actually? Has just never been decided. The Supreme Court has just danced around that issue. Maybe it's the math major in me, but I always want an underpinning to go back to, a substantive underpinning, and when it turns out that the substantive underpinning isn't there, I feel a sense of chaos.

Luigi Zingales: I think that's the right feeling in this moment. It is a sense of chaos.

Bethany McLean: Yes. I don't think there's anything particularly nuanced in this, in that even though my point about billionaire spending, I stand by it. I don't think billionaires are uniform in what they want. I hear your point that it prevents things that should be put on the table, and his point that it's the fear of what will happen if you put something on the table that people don't want.

For sure, I guess the other counter to my argument would be the carried interest tax, that that seems to unite billionaires of all other political proclivities, and that actually is important to the United States budget. Whatever, I just refuted myself. Anyway, the existence of this is just serving to exacerbate our lack of trust and our questions about how our democracy really works, and I think it has been just incredibly dangerous. I don't have that much more to say about it. It's not to me like there are two sides to this.

Luigi Zingales: My only objection is it seems a little bit too clever by half. It's obvious to me that expenditures are not independent. This story of the red boxing, when I saw these students presenting this stuff, I started laughing like there was no tomorrow because they coordinate by announcing, "I want this message to be broadcasted." What is more than coordination than this?

It is a fig leaf that nobody wants to basically take out, and maybe that's the only legal way to do it. Of course, he's the expert here, and so I trust him that this is probably right. It's a little bit depressing because it's pretty obvious to any person who's not blind, super PACs are coordinated. They are not independent. The fact that nobody raise an arm and say, "Let's stop this," is pretty sad.

Bethany McLean: Yes. Although I think that that's maybe a comment on our times, but also a comment in more generally on how hard it is to see the forest for the trees. In other words, we all wave our hands at the trees. One of the really interesting things about this episode for me was stepping back to see, "Oh-

[laughter]

Bethany McLean: -there really is a forest here. It's astonishing to me how many times in life it's actually the obvious that gets missed, and that pointing out and questioning that the obvious doesn't exist or doesn't make sense is what the entire question was. I think about it with John Carreyrou's excellent work on Theranos. It wasn't some thing down in the details that it was like, "Maybe this technology doesn't work."

[laughter]

Bethany McLean: It's a slightly different question, but it's shocking how hard it is to get to that basic question, which actually does connect to our discussion about AI loosely. I do think that that is one thing that humanity hopefully will retain and that will become all the more important in the face of AI, which is the ability to get to the right question. AI may be able to answer any question, but you still have to be able to ask the question, and it's often the asking of the question that is where the real creativity and the real work is.

Luigi Zingales: Speaking of, I realized that not only this year was the 25th anniversary of 9/11, but it was also the 25th anniversary of Enron. [laughs]

Bethany McLean: Yes, it was. It was.

Luigi Zingales: I think it was about this time of the year that the House of Cards started to collapse.

Bethany McLean: It was, yes. So crazy that it's been 25 years. [chuckles] Anyway.